Washington: A fresh online spat involving some of the biggest figures in artificial intelligence turned into a social media spectacle after Elon Musk publicly asked his AI chatbot Grok to roast Dario Amodei, the chief executive of Anthropic.

What began as a policy dispute between an AI company and the US government quickly escalated into a viral exchange on the platform X.

The controversy started after Amodei issued an apology related to a leaked internal message about the Donald Trump administration. Musk's prompt to Grok triggered a chain reaction on social media, eventually leading users to turn the joke back on him by asking the chatbot to roast its own creator.

The episode has reignited debates about AI moderation, the role of tech leaders in shaping online discourse and the risks of encouraging unfiltered responses from chatbots.

The dispute traces back to a difficult week for Anthropic, the company behind the Claude family of AI models.

The United States Department of Defense recently labelled the company a 'supply chain risk', effectively preventing defence contractors from working with its technology.

The decision followed disagreements between Anthropic and the US defence establishment regarding the potential use of its AI models in areas such as surveillance, military intelligence and autonomous weapons systems.

Officials reportedly raised concerns about the company's policies and the potential security implications of using its AI models in defence-related operations.

The designation triggered widespread debate within the technology and national security communities, as artificial intelligence increasingly becomes central to military planning and operations.

The situation intensified when an internal message written by Amodei was leaked online.

In the message, the Anthropic CEO reportedly criticised the Trump administration and suggested that the company was facing hostility partly because it had not offered the kind of praise or financial support that some competitors allegedly had.

The leak quickly spread across social media platforms, drawing attention from technology leaders, policymakers and commentators.

Facing growing criticism, Amodei later issued a public apology through a statement published on the company's website.

'It was a difficult day for the company, and I apologise for the tone of the post. It does not reflect my careful or considered views,' he said.

Amodei also clarified that the message had been written several days earlier and did not represent the company's current assessment of the situation.

 BREAKING: Anthropic CEO just did a complete 180 in live interview

Do you regret saying 'dictator-style praise' about President Trump?

Anthropic CEO:
>'I want to completely apologize for this memo'
>'it was among the most disorienting times in Anthropic's history'
>'i… pic.twitter.com/MTodBFhebU

The Anthropic chief also denied leaking the message himself and emphasised that the company remains committed to supporting US national security initiatives.

'Our most important priority right now is making sure that our warfighters and national security experts are not deprived of important tools in the middle of major combat operations,' he said.

In an interview with The Economist, Amodei elaborated further on the context of the internal communication.

He explained that messages shared on the company's internal Slack channels are often informal and not intended to represent carefully prepared public statements.

'I post things in Slack. I post them a lot. The culture within the company is that I'm very free… it's not what I would say on reflection,' he said.

The interview, widely circulated on X, framed his comments as a dramatic reversal and brought renewed attention to the controversy.

The story took another twist when users on X began experimenting with similar prompts.

Inspired by Musk's instruction, some users asked Grok to roast other public figures using similar wording.

One user eventually turned the spotlight back on Musk by asking the chatbot to generate a roast about him.

Grok responded with an equally harsh and sarcastic message criticising Musk's leadership of companies such as Tesla, SpaceX and Neuralink.

The response quickly went viral, with many users pointing out the irony that Musk's own AI had roasted him using the same style of prompt he had requested for Amodei.

The episode became one of the most widely shared discussions on the platform, blending technology, humour and criticism in a single viral moment.

The incident has sparked a broader conversation about AI governance and moderation.

Experts say the episode illustrates how quickly AI-generated responses can escalate once they are amplified through social media.

While some users praised Grok's unfiltered personality, others warned that encouraging AI systems to produce abusive or extreme content could lead to harmful consequences.

The controversy also highlights the ongoing challenge for AI developers in balancing creative freedom with responsible behaviour.

Meanwhile, Anthropic's conflict with the Pentagon remains unresolved.

The company has said it plans to challenge the 'supply chain risk' designation in court, arguing that the decision could harm innovation and restrict access to advanced AI technologies.

Defence officials, however, have indicated that the restriction will remain in place for now while security concerns are reviewed.

The viral exchange between Musk, Grok and Amodei illustrates how quickly serious policy debates in the technology sector can transform into social media drama. What began as a dispute over national security and artificial intelligence policy soon evolved into a global online spectacle.

As AI tools become increasingly influential in public conversations, incidents like this are likely to intensify discussions about accountability, moderation and the role of technology leaders in shaping the digital environment.